Time： 10:30-12:00 am , May. 28,2012 (January)
Place：Room 446, ICT. CAS
Speaker: Dr. Zhen Fang
This talk consists of two parts, one on processor cache design and the other on SoC applications. In the first half of the talk, I will talk about a high-performance L1 cache design that exploits software semantics for the purpose of saving power. In the second half, I will discuss how we optimized one of the best-known speech recognition engines for the purpose of enabling on smartphone-class devices.
Zhen Fang is a Senior Memory Architect at Nvidia working on GPU and SoC memory subsystem design and verification. Previously he worked at Intel on various projects including high-performance processor core and uncore design, low-cost processor design, and SoC architectures and applications.
Zhen Fang received his Ph.D degree from the University of Utah, and MS and BS degrees from Fudan University. He received a number of awards at Intel for contributions to the Nehalem processor, an Intel-confidential embedded processor, and SoC applications. He has over 20 issued and pending US patents.